DE10029579B4 - Method for orienting the load in crane installations - Google Patents

Abstract

Method for orienting the load in crane installations (1), in which the load suspended on ropes is moved by a specific absolute angle (γ) with a slewing mechanism arranged between the lower bracket (4) arranged on the cable suspension (2) and the load receiving means (3) is arranged and used for rotating the load, characterized in that a control for the slewing suppresses torsional vibrations of the load, as input the absolute rotational angular velocity (γ.) Via a gyroscope sensor is measured, the angular position (c) of the slewing gear and the rope length ( l _S ) are measured and returned to the control input.

Description

The invention relates to a method for orienting the load in cranes according to the preamble of claim 1.

To ensure efficient material flow, most cranes are equipped with a special load handling attachment on the bottom strap of the load rope. The load handling means varies depending on the goods to be transported. For example, a container spreader serves as a load handling device for containers. As far as the cargo is an asymmetrical object, an orientation of the load at the destination point is required. Orientation means that the load at the target point is rotated by a defined angle. For this purpose, a slewing gear is installed in the load receiving means between the cable suspension point and the gripping device for the load.

Now, if such a slewing actuated so can be caused by too fast rotation of the load a torsional vibration, which can be damped by a skilled crane operator by a targeted countermovement of the slewing gear. It depends on the experience and the skill of the crane operator, how quickly he can compensate for such a torsional vibration. For example, with appropriate wind load, a corresponding torsional vibration can also be excited from the outside. These superimposed torsional vibrations are very difficult to be compensated by the crane operator.

Methods are already known for the suppression of pendulum vibrations in load cranes.

That's how it describes DE 127 80 79 A an arrangement for the automatic suppression of oscillations of a suspended by means of a rope at a movable level rope suspension point load on movement of Seilaufhängepunktes in at least one horizontal coordinate, wherein the speed of Seilaufhängepunktes in the horizontal plane by a control loop as a function of one of the Auslenkwinkel of the load rope against the Endlot derived size is influenced.

The DE 20 22 745 A shows an arrangement for the suppression of pendulum vibrations of a load suspended by means of a rope on the cat of a crane, the drive is equipped with a speed device and a Wegregeleinrichtung, with a control arrangement, the cat taking into account the period of oscillation during a first part of so accelerated and retarded during a last part of this path, that the movement of the cat and the vibration of the load at the destination immediately become zero.

From the DE 321 04 50 A1 a device has been known on hoists for the automatic control of the movement of the load carrier with reassurance of the pendulum of the load occurring during acceleration or deceleration of the load hanging on it during an acceleration or deceleration time interval. The basic idea is based on the simple mathematical pendulum. The cat and load mass is not included in the calculation of the movement. Coulombic and velocity-proportional friction of the cat or bridge drives are not considered.

To be able to transport a load body as quickly as possible from the location to the destination, the DE 322 83 02 A1 to control the speed of the drive motor of the trolley by means of a computer so that the trolley and the load carrier are moved during the steady drive at the same speed and the pendulum damping is achieved in the shortest possible time. The from the DE 322 83 02 A1 known computer works according to a computer program to solve the differential equations applicable to the undamped two-mass vibration system formed from trolley and load body, the coulomb and speed-proportional friction of the cat or bridge drives are not taken into account.

In the from the DE 37 10 492 A1 have become known methods, the speed between the destinations on the way chosen so that after covering half the total distance between the starting place and destination of the pendulum deflection is always equal to zero.

That from the DE 39 33 527 A1 Known method for damping of load oscillations comprises a normal speed position control.

The DE 691 19 913 T2 deals with a method for controlling the adjustment of a swinging load, in which in a first control loop the deviation between the theoretical and the actual position of the load is formed. This is derived, multiplied by a correction factor and applied to the theoretical Position of the mobile vehicle added. In a second control loop, the theoretical position of the mobile carrier is compared with the actual position, multiplied by a constant and added to the theoretical speed of the mobile carrier.

The from the DE 127 80 79 A . DE 393 35 27 A1 and DE 691 19 913 T2 have become known control method need for load swing damping a cable angle sensor. In the extended version according to the DE 44 02 563 this sensor is also required. Since this rope angle sensor causes considerable costs, it is advantageous if the load oscillation can be compensated without this sensor.

In the DE 20 22 745 A Several sensors are required for load swing damping. So must in the DE 20 22 745 A at least one speed and position measurement of the trolley are made.

Also the DE 37 10 492 A1 requires at least the cat or bridge position as an additional sensor.

As an alternative to this method proposes another approach, for example, from the DE 32 10 450 A1 and the DE 322 83 02 A1 is known to solve the system underlying differential equations and based on a control strategy for the system to determine to suppress a load swing, in the case of DE 32 10 450 A1 the rope length and in the case of DE 322 83 02 A1 the rope length and load mass is measured. In these systems, however, the frictional effects of static friction and velocity-proportional friction, which are not negligible in the crane system, are not taken into account.

In the not pre-published DE 199 20 431 A1 the applicant of the present invention, a method for load oscillation damping on cranes has been implemented with a control algorithm based on the idea that not only the function of the target load position as a function of time are generated as a guide variables, but also the function for the Target load speed, target load acceleration, the target load jerk and the derivative of the target load jerk and weighted in a pilot block so weighted be switched to the crane system, that the resulting overall system of Krandynamik and feedforward speed distribution, acceleration faithful, faithful and faithful to the derivation of the Ruckes works. As minimum input variables for this prioritized but not previously published method, the cable length and the load mass are needed.

None of the methods described above is concerned with the initially described problem of torsional vibrations in the operation of the slewing gear.

From the WO 97/19888 A1 is known a crane system in which the loads are provided with a four-rope suspension. Here, to prevent torsional movement of the load, at least one of the ropes of the four-wire suspension is influenced.

The DE 19907989 A1 describes a method for the path control of cranes with a single-wire suspension, in which torsional vibrations can be damped about a rotational vector aligned in the direction of the X-coordinate

The object of the present invention is therefore to provide a method for orienting the load in cranes according to the preamble of claim 1, with which a load without excitation of torsional vibrations can be rotated to a defined angular position and are effectively damped with the possibly externally excited torsional vibrations can.

According to the invention the object is achieved by a method with the feature combination of claim 1. Here, a control of the slewing gear is realized, which is based on the measurement of the absolute rotational angular velocity by means of a gyroscope sensor and the angular position of the axis of rotation of the slewing gear.

Further details and advantages of the invention will become apparent from the subsequent claims to the main claim.

Thus, the rotational movement of the load and the gripping device for the load can be detected with a gyroscopic sensor. Since the measurement signal is partially noisy with available gyroscope sensors and is corrupted by drift and offset, according to a further advantageous embodiment of the invention, the offset is estimated and compensated in a so-called interference observer module. An observer computes the absolute angular position of the load based on the idealized dynamic model of the array of the sensor signal from the gyroscope sensor.

In the control according to the invention, a control algorithm can advantageously be used in which the time functions for the setpoint position, the setpoint speed, the setpoint acceleration, the setpoint jerk and the derivative of the setpoint jerk form in a so-called path planning module. These functions are so weighted in a feedforward block applied to the crane system, that the resulting overall system of Krandynamik and precontrol speed, true to the faithful, faithful and faithful to the derivative of the jerk works. In this model, the rope length and the load mass are considered as additional variable parameters.

Further details and advantages of the invention will be explained in more detail with reference to an embodiment shown in the drawing. Show it:

1 the basic structure of a crane with load handling equipment

2 : the cable suspension of the control and rotary axis on the load handling device

3 , the overall structure of the controller

4 : exemplary time functions of the path planning module

5 : the structure of the axis controller

6 : the structure of the state governor

7 : the structure of the pilot control

8th : the structure of the observer

In 1 is the basic structure of a crane 1 with a load handling device 3 shown. Between load handling devices 3 and bottom bottle 4 the rope suspension 2 is a rotation axis 5 arranged around which the bottom block of the cable suspension with respect to the actual load-carrying means can be rotated by a motor. This allows the load to be rotated by the angle γ.

Based on 2 Now a dynamic model is derived to describe this process. The essential effect in the orientation of the load is based on the fact that with the axis of rotation of the lower bottle 4 the rope suspension 2 opposite the load handler 3 is twisted. The position of the rotation axis corresponds to the variable c. As a result, the four ropes running up to the crane trundle 21 against the direction of rotation of the axis of rotation. The Verdrilling corresponds to the difference angle γ _drill = γ - c (1)

This leads to a slight lifting of the load. The diagonal distance between the suspension cables is d _c . By _{Verdrilling the suspension cables are} deflected by the angle φ _1drill .

l _S here corresponds to the length of the suspension ropes 21 between hoist winch and bottom block 4 ,

angehoben. Damit entsteht ein rückwirkendes Drehmoment

mit der beschleunigenden Kraft F_drill = m_Lgsinφ_1drill (4), wobei m_L die Masse der Last ist.This will relieve the load

raised. This creates a retroactive torque

with the accelerating force F _drill = m _L gsinφ _1drill (4), where m _{L is} the mass of the load.

The torque M _drill is reacted in an opposite rotational movement. The result is a torsional vibration, which is described by the following differential equation. (Θ _Lc + Θ _Uc ) γ .. _drill = -M _drill - M _c (5)

Θ _Lc is the moment of inertia when the effector rotates about the axis of rotation, Θ _Uc is the moment of inertia when the lower block rotates about the axis of rotation, M _c is the reaction of the driving torque of the axis of rotation to the twist angle γ _drill . Depending on the acceleration of the axis of rotation is the driving moment M _c = Θ _Lc c .. (6)

Eq. 4 is now linearized by sinφ ≈ φ _1drill . This gives the following equation of motion

Eingangsmatrix:

Systemmatrix:

In order to design a regulator which suppresses the torsional vibrations which inevitably occur when the load is rotated, the differential equation Eq. 7 transferred to the state space representation. As state variables, the twist angle, the angular position of the axis of rotation and their derivatives are defined. This gives the following state space model: ẋ _c = A _c x _c + B _c u _c y _c = C _c x _c y _mc = C _mc x _c (8) with state vector:

Input matrix:

Matrix:

Input vector:

• u _c = c .. = c .. _soll

Output matrix of the controlled variable:

• C _c = [1 1 0 0]

Output vector of the controlled variable:

• y _c = γ

Output matrix of the measured quantity:

Output vector of the measured quantities:

The dynamics of the drive unit of the rotation axis is neglected. Thus, as the input vector of the system instead of the target acceleration of the axis of rotation, the acceleration of the axis of rotation can be used. The input vector of the system description is also the output variable of the subsequently derived controller.

The measured variables are the absolute rotational angular velocity and the angular position of the rotational axis. The rotational angular velocity is detected by a gyroscope sensor. Since its measured value is corrupted by drift and offset, a disturbance observer must support the measurement data evaluation. The position of the rotation axis is detected with an absolute encoder. The rotational angular velocity of the rotation axis is formed by real differentiation.

For the following design of precontrol and state control, the model representation according to Eq. 8 and Eq. 9 extended by the connection of the reference variable vector w _c , via the precontrol matrix S _c and the state feedback via the regulator matrix K _c . This gives you u _c = S _c · w _c - K _c · x _c (10) where reference variable vector:

Pilot control matrix:

• S _c = [K _Vc0 K _Vc1 K _Vc2 K _Vc3 K _Vc4 ] (11)

Knob Matrix:

• K _c = [k _c1 k _c2 k _c3 k _c4 ]
• in which c .. _{soll, back} = - K _c x _c and c .. _{soll, vorst} = S _c w _c

In summary, the following overall structure of the control of the axis of rotation can be represented ( 3 ). The operator becomes a target position γ _target, for example, via the host computer 36 or a target speed γ. _aim for example via the radio remote control 35 specified. In the railway planning module 31 From this, the reference time functions for the setpoint position γ _Lref , the setpoint speed, are _calculated γ. _{L ref} , the target acceleration γ .. _ref , the target jerk γ ... _Lref and the derivative of the target jerk formed γ ^(V) _Lref, wherein the kinematic constraints such as the maximum velocity ν _max, the maximum acceleration a _max and the maximum jerk j _max are always met. In 4 exemplary generated reference time functions are shown as they are for a similar system in the DE 199 20 431 A1 already explained. The reference time functions are the output variables of the path planning module 31 and at the same time the input variables for the axis controller module 33 whose structure is in 5 is shown in more detail.

The axis controller module consists of the pilot control module 51 , the state controller module 53 and the observer module 55 , Input variables are the reference time functions from the path planning module, output variable is the nominal acceleration of the axis of rotation c .. _should , Required measured quantities are the rope length l _S , the load mass m _L , the position of the rotation axis c and the absolute angular velocity of the lifting device γ ..

The following are the modules 51 . 53 and 55 explained in more detail.

The state controller 53 for the rotation axis is designed according to the Polvorgabeverfahren. The characteristic equation of the system with state controller is det (s I - A _c + B _c · K _c ) = 0 (12)

vorgegeben. Die r_ci sind so zu wählen, daß das System stabil ist, die Regelung hinreichend schnell bei guter Dämpfung arbeitet und die Stellgrößenbeschränkung bei typischen auftretenden Regelabweichungen nicht erreicht wird. Werden die Gleichungen Gl. 12 und Gl. 13 gleichgesetzt, so ergeben sich die zu bestimmenden Reglerverstärkungen k_c1 bis k_c4 zu

The desired dynamics of the controlled system is via the Poynom

specified. The r _ci should be chosen so that the system is stable, the control operates sufficiently fast with good damping and the manipulated variable limitation is not reached for typical occurring control deviations. If the equations Eq. 12 and Eq. 13, the controller gains k _c1 to k _c4 to be determined result

Dependent system parameters in the controller gains k _c1 to k _c4 are the variables of the load mass m _L , the diagonal distance of the suspension cables d _c , the cable length l _S , the moment of inertia when rotating about the vertical axis for the load-carrying device Θ _Lc , and the lower block Θ _Uc . Of these, the variables m _L , l _S , Θ _{Lc are} variable. The cable length l _S and the load mass m _L are available as measured variables. This can do that Moment of inertia Θ _Lc , assuming homogeneous mass distribution are approximately derived from the load mass m _L on the geometric dimensions of the grid box. As a result, the moment of inertia can also be attributed to the change in the load mass. The variable parameters in the adaptive tracking of the controller gains are thus the load mass m _L and the rope length l _S. The structure of the state controller module is again in 6 shown. The state variables of the twist angle γ _drill and its derivative, which from the rotational speed γ. and the position of the axis of rotation c are determined, and the position of the axis of rotation c itself and its derivative are returned to the control input via the controller gains k _c1 to k _c4 . The proportion of the manipulated variable determined by the feedback is called c .. _soll.rück designated.

The following is the design of the pilot module 51 to be shown. The railway planning module 31 generates the reference time functions γ _{Lref of} the target angular position, angular velocity, acceleration, and the jerk for the orientation γ of the load in the working space. These are interpreted by the axis controller module for the rotation axis as a reference variable vector w _c , which is given to the input u _c via the feedforward matrix.

hergeleitet. Die Auswertung von Gl. 15 führt auf eine Übertragungsfunktion mit Nennergrad entsprechend der Systemordnung von n = 4.First, the transfer function

derived. The evaluation of Eq. 15 leads to a transfer function with denominator degree according to the system order of n = 4.

Due to the denominator degree 4 of Eq. 16 is to provide a lock up to grade 4. For the precontrol itself therefore results after evaluation of Eq. 10 and 11 and transformation in the frequency domain the following transmission behavior.

This gives the following total transfer function:

For the calculation of the gains K _V0 to K _V4 , due to the degree 4 of the denominator polynomial in Eq. 16 only the coefficients b ₄ to b ₀ and a ₄ to a ₀ of interest. Ideal system behavior with regard to position, speed, acceleration, jerk and, if necessary, the derivative of the jerk is obtained precisely if the transfer function of the overall system consisting of precontrol and transfer function satisfies the following conditions in its coefficients b _i and a _i :

After evaluation analogous to Eq. 7-17 is thus obtained for the pilot control gains

The expressions of Eq. 20 show that the adaptive tracking of the gains in the _precontrol must take into account the system parameters m _L , d _c , l _S , Θ _Lc and Θ _Uc . As with the state controller module, homogeneous mass distribution is assumed and the moment of inertia Θ _{Lc is} calculated approximately from the load mass and the geometric dimensions of the grid box. The variable parameters in the adaptive tracking are thus the load mass m _L and the cable length l _S. The structure of the feedforward control is in 7 shown. Input variables are the reference time functions from the path planning module, output variable is the proportion of feedforward control c .. _{should, vorst} on the manipulated variable c .. _should ,

To measure the absolute angular velocity of the load, a gyroscope sensor is installed on the load handler. The measurement signal of the sensor is superimposed on the principle of measurement with a significant offset. The offset on the measuring signal causes position errors of the control in the orientation of the load. Therefore, in an observer, the offset error is estimated and compensated. For this, the offset error is the disturbance variable γ. _offset introduced. The disturbance is assumed to be constant as sections. The fault model is accordingly γ .. _Offset = 0 (21)

wobei in Ergänzung zu Gl. 9 die folgenden Matrizen und Vektoren eingeführt werden.The state space representation of the submodel for the Drebachse according to Eq. 8 and Eq. 9 is extended by the fault model. In the present case, a complete observer is derived. The observer equation for the modified state space model is therefore:

in addition to Eq. 9 the following matrices and vectors are introduced.

State vector:

Input matrix:

Matrix:

Störbeobachtermatrix:

Observers output matrix:

For the observer's design, the system of Eq. 23 transformed into the observation normal form. In observational standard form, the observer is designed via pole specification and then the system is transformed back again. The poles r _cz1,2 and r _{cz3,4 are chosen} with a multiplicity of two and the pole r _cz5 with a multiplicity of one. The observer observer matrix for the observer 55 is then

With the representation according to Eq. In turn, there is an analytic expression dependent on the system parameters m _L , d _g , l _S , Θ _Lc . For the adaptation of the Störbeobachter 55 the variables m and l _{L S} are required. The structure of the observer 55 is in 8th shown.

From the measured variables of the position of the axis of rotation c and the rotational speed γ. of the load handling device is the offset error on the observer γ. _offset determined. This makes it possible to measure the measured value of the rotational speed γ. to correct and thus reliably calculate the twist angle γ _drill for the state _controller .

After in the previous the individual submodules 51 . 53 and 55 were introduced, the entire structure should now be compared again 5 be shown to clarify the relationships between the sub-modules again. 5 shows the structure of the axis controller module for the axis of rotation of the lifting device. Input variables for the pilot control module 51 are the reference time functions γ _{Lref of} the _path planning _module 31 , Due to the system order n = 4, a connection can be made up to the derivation of the target jerk. Output is c .. _{should, vorst} , About the state controller 53 become the state variables γ, γ., c, c. on the entrance as c .. _{shall, return} recycled. The measured variables are the position of the axis of rotation c and the speed c formed by real differentiation. and the offset rotational speed γ. in front. To compensate for the offset error in the gyroscope signal is therefore a Störbeobachtermodul 55 introduced that the offset γ. _offset underestimated. Subsequently, the measuring signal of the gyroscope sensor is corrected by this estimated offset prior to switching to the state controller and before its integration for deriving the position signal γ. Therefore, for the function of the state controller module 53 in this case, the observer 55 absolutely necessary. Output variable of the axis controller module is the nominal acceleration of the axis of rotation c .. _should ,

Claims (7)

Method for orienting the load in crane installations ( 1 ), in which the load suspended on ropes is shifted by a certain absolute angle (γ) with a slewing gear which is suspended between the cable suspension ( 2 ) arranged lower bottle ( 4 ) and the load handling device ( 3 ) And,) the angular position is used to rotate the load characterized in that a control for the slewing gear torsional vibrations of the load is suppressed, using as input the absolute rotational angular velocity (γ, a gyroscopic sensor is measured, (c) of the rotating mechanism and the cable length (l _S ) are measured and returned to the control input. Method according to Claim 1, characterized in that the regulation for the slewing gear positions the load at a predetermined desired rotational angle. Method according to one of claims 1 to 2, characterized in that, taking into account the cable length (l _S ) and the load mass (m _L ) in a railway planning module ( 31 ) the time functions are formed at least one of the variables of the desired angular position, desired angular velocity, desired angular acceleration and the desired angular pressure and the derivative of the jerk for the orientation γ of the load in the working space and these in a pilot control block ( 51 ) of an axis controller module ( 33 ) are weighted with feedforward gains K _Vi so that the coefficients of the resulting transfer function from Krandynamik and feedforward of the form

the following conditions are sufficient

Method according to one of claims 1 to 3, characterized in that the solid defined by the transfer function control amplifications as a function of the load mass (m _L) and the cable length (l _S) are calculated. Method according to one of claims 1 to 4, characterized in that the path planning module the time functions of the target position (γ _Lref ), the target speed ( γ. _{L ref} ), the target acceleration ( γ .. _Lref ) and the target jerk ( γ ... _Lref ), taking into account the kinematic constraints. Method according to Claim 4, characterized in that the path planning _{module also} generates the time function for deriving the desired jerk (γ ^(IV) _Lref ). Method according to one of claims 1 to 6, characterized in that an offset occurring in the measuring signal of the gyroscope sensor on the measuring signal in an interference observer module ( 55 ) is eliminated due to estimation and compensation of the offset error.

Images